1

我作为 gensim 教程在 195145 个文档、6636308 个特征、188901082 个非零条目上运行 LAD。代码很简单：

from gensim import corpora, models, similarities
class MyCorpus(object):
    def __iter__(self):
        for line in open('/home/pda/xxz149/LDA/DrugPatents.csv'):
            # assume there's one document per line, tokens separated by ','
            yield dictionary.doc2bow(line.lower().split(','))
dictionary = corpora.Dictionary.load('/home/pda/xxz149/LDA/DrugPatent.dict')
corpus = MyCorpus()
lda = models.ldamodel.LdaModel(corpus, num_topics = 300,id2word=dictionary,distributed = False,chunksize = 15, passes = 1 )
lda.save('/home/pda/xxz149/LDA/lda_DrugPatent.model')

但我遇到了价值错误：

  File "/usr/lib/python2.6/site-packages/gensim-0.10.0rc1-py2.6.egg/gensim/models/ldamodel.py", line 79, in __init__

    self.sstats = numpy.zeros(shape)
ValueError: array is too big.

gensim 是内存友好的，为什么会发生这种情况？我怎样才能通过？

4

0 回答 0